On January 9, IPRdaily, the world’s authoritative intellectual property third-party organization, released the “Brief Report on Patent Analysis of AI Security and Credibility Key Technologies”. The report cites the international authoritative database Derwent World Index. As of November 28, 2022, the number of applications for a total of 17,605 patent families in the global AI security and trustworthy technology field is 9,567 more than in 2021. IBM, NTT, Tencent, Ant and many other companies rank in the forefront. Among them, Alipay, a subsidiary of Ant Group, ranks first in the world in the number of patent applications and authorizations in this field. This is also the number of AI security and trustworthy technology patents of Ant Group for two consecutive years The world’s first.
Top 10 Authorized Numbers of Derwent Patent Families for AI Safe and Trustworthy Technology (Source: IPRdaily Chinese Website)
Top 10 Derwent Patent Families of AI Safe and Trustworthy Technology (Source: IPRdaily Chinese Website)
It is worth mentioning that 62% of patent innovations in this field originate in China, which is the region most concerned about AI security and trustworthy technology and innovative research in the world. The report shows that 15 of the top 20 innovators in the number of patent families are Chinese enterprises, scientific research institutions and colleges.
The report sets that the key technologies for AI security and trustworthiness include four attributes: AI robustness, AI data privacy protection, AI explainability, and AI fairness, which are also the four key features of the industry-consensus trusted AI technology system. Specifically, it includes multiple technologies such as intelligent confrontation, intelligent decision-making, graph computing, privacy computing, and interpretable machine learning.
With the large-scale industrial implementation of AI technology, problems such as “algorithm security”, “data discrimination”, and “privacy leakage” have become prominent, which also points AI to the dilemma of “not being trusted”. The development of trusted AI and the use of technology to promote artificial intelligence to be transparent, explainable, safe and controllable, and diverse and inclusive have become the direction that industry, academia, and regulatory agencies are promoting. In the past few years, technology companies such as Microsoft, Google, IBM, Tencent, and Ant have reinvested in trusted AI, and a wide range of industries have landed. In the “White Paper on Artificial Intelligence (2022)” issued by the Institute of Information and Communications Technology, “responsible and credible artificial intelligence” has become one of the new “three-dimensional” development coordinates of AI products.
According to public information, Ant Group has been developing trusted AI technology since 2015, which is mainly used in security risk control businesses such as anti-fraud, anti-money laundering, corporate joint risk control, and data privacy protection, and has formed an “IMAGE intelligence technology” based on trusted AI. Risk control system”, which solves many problems in the risk control scene. For example, the “Ant Trusted Privacy Sandbox (AntDTX)” based on trusted privacy computing and terminal-edge-cloud collaborative risk control solutions has upgraded the local risk perception capabilities of over 24 million mobile devices; based on trusted AI and blockchain The “anti-fraud alliance chain” cooperates with the public security, the People’s Bank of China, and operators to solve the problem of risk information sharing, and has shared more than 20 million black and gray production information.
In addition to using it for its own business, Ant Group also opens up its useful and trusted AI technology to the ecology. Its security technology commercialization product “ZOLOZ” has served more than 1,000 customers in more than ten industries such as finance, travel, and e-commerce in China, and has helped dozens of partners overseas to serve more than 100 million customers ; The “Ant Jian AI Security Testing Platform” created has opened AI robustness testing to developers around the world free of charge. In August last year, three trusted AI technology practice solutions of Ant Group, including the “Ant Jian AI Security Detection Platform”, were awarded the “2022 Trusted Artificial Intelligence Practice Benchmarking Case” by the Academy of Information and Communications Technology
Derwent World Patent Index is the world’s most authoritative patent value-added data, used by patent offices in more than 40 countries/regions for official patent searches and examinations. The Derwent World Patent Index groups multiple patent applications on the same technical subject into a single Derwent patent family.
This article is transferred from: https://www.leiphone.com/category/industrynews/2M8LWVqVlDCJxtGc.html
This site is only for collection, and the copyright belongs to the original author.